There is an urgent need for development of chemotherapeutic agents in the therapy of viral diseases. In particular treatment of diseases caused by retroviruses presents one of the most difficult challenges in human medicine. While a number of antiviral agents which are registered or are currently under study can effectively cure disease, relieve symptoms or substantially prolong the intervals among the recurrences of certain chronic viral infections, such positive outcomes have not yet been achieved in many instances, notably that of AIDS, as an example of retroviral disease. Selectivity of antiviral action, which is an important requirement for novel antiviral agents, has not been achieved.
Most of the compounds which are clinically useful for antiviral chemotherapy are nucleosides, modified in either the purine or pyrimidine base and/or the carbohydrate moiety. Such compounds mainly act in processes related to the synthesis of viral nucleic acids; their action depends on ability to undergo phosphorylation and subsequent transformation to the triphosphates. One problem in administering modified nucleosides is the absence of suitable phosphorylating activity in the host cell and the existence of viral strains lacking virus-specific phosphorylating activity. While enzymatically resistant nucleotide analogs might appear to be particularly useful as potential antivirals, their polar character prevents effective entry of these analogs into the cells, as does lack of appropriate nucleotide receptors at the cellular membrane.
This difficulty appears to be overcome in the series of acyclic nucleotide analogs which contain an aliphatic chain, bearing hydroxyl groups, replacing the sugar moiety. For example, the phosphates or phosphonic acid derivatives derived from the antiviral nucleoside analog ganciclovir (Cytovene) are reported to possess an anti-herpes virus activity (Reist at al., in "Nucleotide Analogs as Antiviral Agents", ACS Symposium Series, No. 401, pp. 17-34 (1989); Tolman, ibid, pp. 35-50; Prisbe et al., J Med Chem (1986), 29:671).
The following formulas describe several classes of prior art compounds:
______________________________________ ##STR2## (1) R = CH.sub.2 OH (2) R = H (3) R = CH.sub.2 ______________________________________ F
Another group of antiviral compounds where the antiviral action is less strictly limited by the nature of the heterocyclic base includes phosphonic acid analogs in which a phosphonic acid moiety is linked to the hydroxyl group of an aliphatic chain sugar substitute via a methylene unit. Examples of such compounds are HPMP-derivatives (1) which were disclosed by the UK Patent Application No. 2 134 907 and PV-3017, now published on Dec. 30, 1986 as EP 206,459 of Holy et al. Such compounds act exclusively against DNA viruses as reported by De Clercq et al. in Nature (1986) 323:464-467, and reviewed by Holy et al. in Antiviral Res (1990) 13:295.
A similar type of antivirals is represented by PME-derivatives (2) disclosed by European Patent Application 0 206 459 by Holy et al. and described in detail by De Clercq at al. in Antiviral Res (1987) 8:261, and by Holy et al. in Collection Czech Chem Commun (1987) 52:2801; ibid. (1989), 54:2190). These compounds act against both DNA viruses and retroviruses, including HIV-1 and HIV-2. The adenine derivative, PMEA, was demonstrated to exhibit an outstanding activity against Moloney sarcoma virus in mice, simian immunodeficiency virus in Rhesus monkeys as well as feline immunodeficiency virus in cats (Balzarini et al., AIDS (1991) 5:21; Egberink et al., Proc Natl Acad Sci U.S.A. (1990) 87:3087).
The extensive structure-activity investigation which concentrated on the modification of the side-chain (described by Holy et al. in "Nucleotide Analogs as Antiviral Agents", ACS Symposium Series No.401 (1989), p. 51) did not reveal any additional substantially active antivirals. Replacement of this hydroxyl by fluorine atom resulted in the FPMP-compounds (3) which, in addition to having some anti-DNA-virus activity display a substantial effect on both HIV-1, HIV-2 and murine sarcoma virus (as taught by Holy et al., Czechoslovak Patent Application PV 2047-90 now published on Oct. 30, 1991 as EP 454,127, and by Balzarini et al., Proc Natl Acad Sci U.S.A. (1991) 88:4961).
The racemic mixtures of 9-(2-phosphono-methoxypropyl)adenine and guanine (PMPA and PMPG) were also described by Holy et al., European Patent Appl. 0206459 (PMPA) and Holy et al., Collection Czech Chem Commun (1988) 53:2753 (PMPA), and by U.S. patent application Ser. No. 932,112 (PMPG). PMPA was devoid of any appreciable antiherpetic effect while any antiherpetic activity of PMPG appeared due to its substantial cytotoxicity. The clinical forms of PMPG, and of the related compound, 9-C3-hydroxy-2-(phosphonomethoxy)propyl) guanine (HPMPG) are disclosed in EP application 452935. For these guanine forms, the R-enantiomers consistently gave greater antiviral activity, especially in regard to the retrovirus HIV. There was little difference between R&S enantiomers in antiviral activity with regarding to some DNA viruses. It cannot be predicted whether this pattern of activity would extend to PMP compounds of other than guanine.
Nothing in the above-cited references or their combination permits any prediction that the resolved enantiomers of the present invention would exhibit antiretroviral activity, or what the enantiomers preference would be.